Abstract Understanding the spatial variation of a plant community’s composition in response to key environmental variables is a fundamental task to identify and promote better conservation strategies. In forested wetlands, hydrological regimes have been identified as crucial factors that shape the community’s attributes and composition patterns. Therefore, the objective of this study was to analyze the response of a tropical forested wetland diversity and structural attributes, as well as its species composition to two types of environmental proxies related to flooding and salinity regimes: microtopography and distance to water bodies. The forest was characterized by sampling all the individuals with DBH ≥ 10 cm found inside twenty-five 25 × 50 m plots. Each plot’s microtopographic variables and distance to water bodies were extracted using a LiDAR point cloud. The response of the community’s attributes to the quantified environmental proxies was analyzed using an RDA ordination. Community composition was the only attribute where the ordination was significant, particularly only its first axis. Additional analysis showed that Rhizophora mangle was gradually substituted by Terminalia buceras as the most abundant species along the RDA first axis. This first axis was significantly correlated with two environmental proxies: distance to the lagoon’s edge and standard deviation of ground surface elevation. Moreover, linear regression analysis revealed that the abundance of three species was explained by these environmental proxies. R. mangle abundance showed a negative response to standard deviation of ground surface elevation, while T. buceras and Chrysobalanus icaco showed a positive response to distance to the lagoon’s edge.